An energy-saving system for converting existing buildings into ultra-low energy consumption buildings

Abstract

This invention discloses an energy-saving system for converting existing buildings into ultra-low-energy buildings, belonging to the fields of computer technology and building energy conservation. It includes a data acquisition module, a fingerprint modeling module, a database and control logic construction module, a collaborative simulation engine module, a multi-objective optimization module, a robust stress testing module, and a scheme decision and output module. This invention employs a reverse engineering approach based on historical data to construct a building thermal fingerprint model, coupled with active control logic for collaborative simulation and multi-objective iterative optimization. This enables accurate diagnosis of the building's thermal status, selection of Pareto optimal solutions, and verification of reliability under extreme climate conditions, thereby improving the energy efficiency, temperature comfort, and all-weather operational robustness of the renovation scheme.

Description

Technical Field

[0001] This invention relates to the fields of computer technology and building energy conservation, and in particular to an energy-saving system for converting existing buildings into ultra-low energy consumption buildings. Background Technology

[0002] Energy-saving retrofitting of existing buildings is a crucial step in achieving urban carbon reduction targets. Upgrading existing buildings to ultra-low energy buildings is an important direction for technological development. Ultra-low energy buildings not only require extremely low operating energy consumption but also have stringent standards for maintaining a constant temperature and comfort level indoors. To achieve this goal, the design of retrofitting plans requires precise assessment of the building's current condition and the scientific selection of passive retrofitting measures such as insulation and window / door systems, while also appropriately integrating future high-efficiency active environmental control systems.

[0003] In related technologies, Chinese invention patent CN117314199A discloses a data processing-based building energy-saving renovation assessment system and method, which includes: a monitoring point determination module for determining multiple renovation monitoring points; a data acquisition module for acquiring pre-renovation and post-renovation building data from multiple renovation monitoring points; a scheme evaluation module for predicting and optimizing the renovation effect of the renovation scheme to be evaluated based on the pre-renovation building data; and a renovation assessment module for determining the renovation effect based on the post-renovation data.

[0004] Regarding the aforementioned technologies, the inventors believe that conventional technical solutions, which typically rely on physical models built from original building design drawings, fail to reflect the actual degradation of the building envelope's thermal performance due to material aging, construction defects, and other factors during long-term use. Furthermore, passive retrofitting schemes are often disconnected from the selection and control strategies of active systems, failing to consider their dynamic interactions and synergistic effects. In addition, the evaluation criteria are singular, overemphasizing energy consumption indicators while lacking effective quantitative assessment and verification methods for indoor temperature stability and system reliability under extreme climatic conditions. Therefore, retrofitting schemes generated by existing technologies often suffer from a disconnect between simulation results and the current situation, a mismatch between energy saving and comfort goals, and a lack of all-weather operational robustness, making it difficult to meet the stringent requirements of ultra-low energy building retrofitting. Summary of the Invention

[0005] To address the aforementioned issues, this invention provides an energy-saving system for converting existing buildings into ultra-low-energy buildings. It employs a data-driven fingerprint modeling approach combined with multi-objective collaborative simulation optimization and robustness testing to accurately and efficiently generate a final renovation plan that balances energy efficiency, comfort, and high reliability.

[0006] To achieve the above objectives, this application adopts the following technical solution:

[0007] Firstly, an energy-saving system for retrofitting existing buildings into ultra-low-energy buildings is provided, comprising: a data acquisition module for acquiring a historical operating dataset containing historical outdoor meteorological data sequences, historical indoor temperature data sequences, and historical equipment operating status data sequences; a fingerprint modeling module for constructing a building thermal fingerprint model characterizing the building's actual thermal dynamic response based on the historical operating dataset using a system identification algorithm; a database and control logic construction module for establishing a database of retrofit measures containing several energy-saving retrofit components and establishing an active environmental system control logic model containing frequency conversion regulation strategies and intelligent defrosting trigger strategies; and a collaborative simulation engine module for selecting a subset of retrofit measure parameters from the retrofit measure database and coupling the building thermal fingerprint model, the subset of retrofit measure parameters, and the active environmental system control logic model. The system executes a co-simulation to generate a comprehensive performance evaluation result corresponding to the subset of parameters for the modification measures. A multi-objective optimization module uses a multi-objective optimization algorithm to iteratively optimize different subsets of modification measure parameters, aiming to optimize the predicted energy consumption and predicted temperature stability indices in the comprehensive performance evaluation result, and selects a Pareto optimal solution set. A robustness stress testing module acquires extreme climate test sequences for simulating extreme weather, performs robustness stress tests on each solution in the Pareto optimal solution set, and obtains robustness evaluation indices for each solution. A solution decision and output module determines the final modification solution from the Pareto optimal solution set based on the comprehensive performance evaluation result and the robustness evaluation indices, and outputs modification decision data corresponding to the final modification solution, including three-dimensional component attributes and system control parameters.

[0008] Based on the above technical solution, in the energy-saving system for converting existing buildings into ultra-low energy consumption buildings provided in this application, the method of data-driven fingerprint modeling combined with multi-objective collaborative simulation optimization and robustness testing can accurately and efficiently generate the final renovation plan that takes into account energy saving, comfort and high reliability.

[0009] In conjunction with the first aspect above, in one possible implementation, the fingerprint modeling module includes: a data preprocessing unit, used to perform time alignment and data cleaning on the historical operating dataset to generate a standardized time-series dataset; a variable definition unit, used to input the standardized time-series dataset into the system identification algorithm, using the historical outdoor meteorological data sequence and the historical equipment operating status data sequence as system input, and the historical indoor temperature data sequence as system output; and a parameter fitting unit, used to iteratively calculate and fit an internal parameter set that minimizes the error between the system simulation output and the system output through the system identification algorithm, and use the internal parameter set to construct a building thermal fingerprint model.

[0010] In conjunction with the first aspect above, in one possible implementation, the collaborative simulation engine module includes: a model update unit, used to update the physical attribute parameters in the building thermal fingerprint model according to the subset of parameters of the renovation measures, and generate a hypothetical post-renovation building model; a logic loading unit, used to load the active environmental system control logic model into the hypothetical post-renovation building model, forming an integrated digital twin simulation; a dynamic simulation unit, used to drive the integrated digital twin simulation to run under typical meteorological data, and record the cumulative system energy consumption value and indoor temperature fluctuation data during the simulation process; and an index calculation unit, used to calculate the predicted energy consumption index based on the cumulative system energy consumption value, and calculate the predicted temperature stability index based on the indoor temperature fluctuation data, which together constitute a comprehensive performance evaluation result.

[0011] In conjunction with the first aspect above, in one possible implementation, the database and control logic construction module includes: a frost data processing unit, used to acquire the historical operating dataset of the outdoor unit of the active environmental system, and process the dataset to generate a multi-dimensional environmental parameter sample containing frost event markers; a prediction model training unit, used to train a frost event prediction model based on the multi-dimensional environmental parameter sample to predict frost risk according to real-time environmental parameters; and a control command integration unit, used to integrate the frost event prediction model into the active environmental system control logic model, so as to generate a pre-defrosting control command when the frost risk output by the frost event prediction model exceeds a preset trigger threshold.

[0012] In conjunction with the first aspect above, in one possible implementation, the multi-objective optimization module includes: a population initialization unit, used to initialize a population composed of several subsets of modification measure parameters; a fitness evaluation unit, used to call the co-simulation engine module to perform co-simulation steps for each subset of modification measure parameters in the population, obtain the corresponding comprehensive performance evaluation result, and calculate the fitness based on the comprehensive performance evaluation result; an evolutionary operation unit, used to generate a new generation population through selection, crossover, and mutation operations according to the fitness; and a loop iteration unit, used to repeatedly trigger the execution steps of the fitness evaluation unit and the evolutionary operation unit until a preset termination condition is met, and output the set of non-dominated solutions in the final generation population as a Pareto optimal solution set.

[0013] In conjunction with the first aspect above, in one possible implementation, the robustness stress testing module includes: an extreme operating condition operation unit, used to place the integrated digital twin simulation corresponding to each scheme into an extreme climate test sequence for operation; a failure event monitoring unit, used to monitor and record temperature exceedance events and defrost temperature drop events occurring in the integrated digital twin simulation under the extreme climate test sequence; and a reliability quantification unit, used to calculate and generate robustness evaluation indicators characterizing the reliability of the scheme under extreme operating conditions based on the quantified data of the temperature exceedance events and defrost temperature drop events.

[0014] In conjunction with the first aspect above, in one possible implementation, the index calculation unit includes: a deviation extraction subunit, used to extract a temperature deviation sequence from the indoor temperature fluctuation data; a sub-index calculation subunit, used to calculate the standard deviation of the temperature deviation sequence as a first stability sub-index, and calculate the percentage of time the indoor temperature exceeds a preset range as a second stability sub-index; and a weighted fusion subunit, used to perform weighted fusion of the first stability sub-index and the second stability sub-index to generate a predicted temperature stability index.

[0015] In conjunction with the first aspect above, in one possible implementation, the scheme decision and output module includes: a matrix construction unit, used to standardize the predicted energy consumption index, predicted temperature stability index, and robustness evaluation index of each scheme in the Pareto optimal scheme set to generate a standardized performance matrix; a weighted scoring unit, used to assign weight coefficients to each performance index in the standardized performance matrix and calculate the comprehensive score of each scheme; and a scheme selection unit, used to select the scheme with the highest comprehensive score as the final modification scheme.

[0016] In conjunction with the first aspect above, in one possible implementation, the scheme decision and output module further includes: a model-driven update unit, used to import the three-dimensional component attributes into an associated building information model platform to update the physical attribute parameters of the corresponding model component in the platform; and a configuration file generation unit, used to encapsulate the system control parameters into a configuration file that can be read by the controller and send it to the associated active environmental system controller.

[0017] Secondly, a method for converting existing buildings into ultra-low energy consumption buildings is provided, comprising: acquiring a historical operation dataset containing historical outdoor meteorological data sequences, historical indoor temperature data sequences, and historical equipment operating status data sequences; based on the historical operation dataset, constructing a building thermal fingerprint model characterizing the building's true thermal dynamic response through a system identification algorithm; establishing a database of renovation measures containing several types of energy-saving renovation components, and establishing an active environmental system control logic model containing frequency conversion regulation strategies and intelligent defrosting triggering strategies; selecting a subset of renovation measure parameters from the renovation measure database, coupling the building thermal fingerprint model, the subset of renovation measure parameters, and the active environmental system control logic model, and running a collaborative simulation... If true, generate a comprehensive performance evaluation result corresponding to the subset of parameters of the modification measures; adopt a multi-objective optimization algorithm to iteratively optimize different subsets of modification measure parameters with the goal of optimizing the predicted energy consumption index and predicted temperature stability index in the comprehensive performance evaluation result, and select a Pareto optimal solution set; obtain an extreme climate test sequence for simulating extreme weather, and conduct robustness stress tests on each solution in the Pareto optimal solution set to obtain the robustness evaluation index of each solution; based on the comprehensive performance evaluation result and the robustness evaluation index, determine the final modification solution from the Pareto optimal solution set, and output the modification decision data corresponding to the final modification solution, which includes three-dimensional component attributes and system control parameters.

[0018] Compared with the prior art, the present invention has the following advantages:

[0019] This invention generates a digital model that highly matches the dynamic thermal response of a specific building by constructing a building thermal fingerprint model based on historical operating data. Compared to traditional modeling methods that rely on general parameter libraries and theoretical calculations, this data-driven approach can more accurately reflect the actual performance of a building, improve the accuracy of simulation predictions for subsequent modification measures, and effectively avoid the performance gap between theoretical design and actual operating results.

[0020] This invention integrates a collaborative simulation engine and a multi-objective optimization algorithm, enabling automated iterative optimization of a subset of parameters for massive renovation measures. It not only evaluates the effectiveness of passive renovation measures but also couples them with the control logic of active environmental systems, achieving a comprehensive consideration of the synergistic benefits between the building envelope and electromechanical systems. This allows for the discovery of a set of Pareto solutions that achieve a globally optimal balance among multiple conflicting objectives such as energy consumption and comfort, thus enhancing the comprehensiveness and scientific rigor of decision-making.

[0021] This invention innovatively introduces a robust stress testing module, expanding the decision-making dimension from performance evaluation under normal operating conditions to reliability verification under extreme climate conditions. By monitoring and quantitatively evaluating failure events of Pareto optimal solutions under extreme conditions, it is possible to select renovation schemes that are not only energy-efficient and comfortable, but also highly resilient and robust under future climate uncertainties, thereby enhancing the long-term operational reliability and climate adaptability of the renovated building.

[0022] This invention achieves seamless data integration from decision analysis to engineering implementation. The final output of renovation decision data includes 3D component attributes that can be directly imported into the Building Information Modeling (BIM) platform, as well as system control parameter configuration files that can be directly sent to the controller. This model-driven automated data flow reduces human error in the information transmission process, improves design and debugging efficiency, and ensures that optimized decisions can be implemented in the physical world with high fidelity.

[0023] It should be understood that the descriptions of technical features, technical solutions, beneficial effects, or similar language in this application do not imply that all features and advantages can be achieved in any single embodiment. Rather, it is understood that the description of a feature or beneficial effect means that a specific technical feature, technical solution, or beneficial effect is included in at least one embodiment. Therefore, the descriptions of technical features, technical solutions, or beneficial effects in this specification do not necessarily refer to the same embodiment. Furthermore, the technical features, technical solutions, and beneficial effects described in this embodiment can be combined in any suitable manner. Those skilled in the art will understand that embodiments can be implemented without one or more specific technical features, technical solutions, or beneficial effects of a particular embodiment. In other embodiments, additional technical features and beneficial effects may be identified in specific embodiments that do not embody all embodiments. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 A flowchart illustrating an energy-saving method for converting an existing building into an ultra-low energy consumption building, as provided in an embodiment of this application;

[0026] Figure 2 A structural architecture diagram of an energy-saving system for converting an existing building into an ultra-low energy consumption building, provided in an embodiment of this application;

[0027] Figure 3This is a Pareto front scatter plot provided in the embodiments of this application.

[0028] Figure 4 This is a dynamic response curve of the robustness stress test provided in the embodiments of this application. Detailed Implementation

[0029] It should be noted that, in this application, the terms "exemplary" or "for example" are used to indicate that something is being described as an example, illustration, or illustration. Any embodiment or design described as "exemplary" or "for example" in this application should not be construed as being more preferred or advantageous than other embodiments or design solutions. Specifically, the use of terms such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0030] The energy-saving system for converting existing buildings into ultra-low energy consumption buildings provided in this application embodiment can be applied to, for example... Figure 1 In one energy-saving method for converting an existing building into an ultra-low energy building, as shown, Figure 1 As shown, the method includes:

[0031] Obtain a historical operational dataset containing historical outdoor meteorological data sequences, historical indoor temperature data sequences, and historical equipment operating status data sequences;

[0032] Based on the historical operating dataset, a building thermal fingerprint model representing the real thermal dynamic response of a building is constructed using a system identification algorithm.

[0033] Establish a database of retrofit measures containing several types of energy-saving retrofit components, and establish an active environmental system control logic model that includes frequency conversion regulation strategy and intelligent defrosting trigger strategy;

[0034] A subset of retrofit measure parameters is selected from the retrofit measure database, and the building thermal fingerprint model, the subset of retrofit measure parameters, and the active environmental system control logic model are coupled together. Co-simulation is then performed to generate a comprehensive performance evaluation result corresponding to the subset of retrofit measure parameters.

[0035] A multi-objective optimization algorithm is adopted to optimize the predicted energy consumption index and predicted temperature stability index in the comprehensive performance evaluation results. Different subsets of modification measures are iteratively optimized to select the Pareto optimal solution set.

[0036] Obtain extreme climate test sequences for simulating extreme weather, conduct robust stress tests on each scheme in the Pareto optimal scheme set, and obtain robustness evaluation indicators for each scheme.

[0037] Based on the comprehensive performance evaluation results and the robustness evaluation index, the final modification scheme is determined from the Pareto optimal scheme set, and the modification decision data corresponding to the final modification scheme, including three-dimensional component attributes and system control parameters, is output.

[0038] like Figure 2 As shown in the figure, this application provides an energy-saving system for converting existing buildings into ultra-low energy consumption buildings, including:

[0039] The data acquisition module is used to acquire historical operation datasets that include historical outdoor meteorological data sequences, historical indoor temperature data sequences, and historical equipment operation status data sequences.

[0040] The fingerprint modeling module is used to construct a building thermal fingerprint model that represents the actual thermal dynamic response of a building based on the historical operating dataset and through a system identification algorithm.

[0041] The database and control logic building module is used to establish a database of renovation measures containing several energy-saving renovation components, and to establish an active environmental system control logic model containing frequency conversion regulation strategy and intelligent defrosting trigger strategy.

[0042] The collaborative simulation engine module is used to select a subset of retrofit measure parameters from the retrofit measure database, couple the building thermal fingerprint model, the subset of retrofit measure parameters, and the active environmental system control logic model, run collaborative simulation, and generate a comprehensive performance evaluation result corresponding to the subset of retrofit measure parameters.

[0043] The multi-objective optimization module is used to employ a multi-objective optimization algorithm to iteratively optimize different subsets of modification measure parameters, with the goal of optimizing the predicted energy consumption index and predicted temperature stability index in the comprehensive performance evaluation results, and to select a Pareto optimal solution set.

[0044] The robustness stress test module is used to acquire extreme climate test sequences for simulating extreme weather, and to perform robustness stress tests on each scheme in the Pareto optimal scheme set to obtain robustness evaluation indicators for each scheme.

[0045] The scheme decision and output module is used to determine the final modification scheme from the Pareto optimal scheme set based on the comprehensive performance evaluation results and the robustness evaluation index, and output the modification decision data corresponding to the final modification scheme, which includes three-dimensional component attributes and system control parameters.

[0046] It is worth noting that a closed-loop decision-making framework based on data-driven and collaborative simulation was constructed. First, a building thermal fingerprint model, capable of accurately reproducing the building's true thermal dynamic response, is generated using historical operational data through a system identification algorithm, serving as a high-fidelity digital twin foundation. Subsequently, this customized model is coupled with a database containing various retrofit components and advanced active system control logic to form an integrated simulation. A multi-objective optimization algorithm drives iterative optimization within this simulation, automatically searching for the Pareto optimal set of solutions that achieve the best balance between energy consumption and comfort from a vast combination of retrofit options. Building upon this, extreme climate stress testing is further introduced to assess the reliability and robustness of these optimal solutions. Finally, a multi-criteria decision-making model integrating performance and robustness selects a unique final retrofit solution from the optimal set and outputs digital decision data that can be directly executed by the design platform and control system.

[0047] In one possible implementation of the embodiments of this application, combined with Figure 2 The fingerprint modeling module includes:

[0048] The data preprocessing unit is used to perform time alignment and data cleaning on the historical running dataset to generate a standardized time-series dataset.

[0049] The variable definition unit is used to input the standardized time series dataset into the system identification algorithm, using the historical outdoor meteorological data sequence and the historical equipment operating status data sequence as system inputs and the historical indoor temperature data sequence as system outputs.

[0050] The parameter fitting unit is used to iteratively calculate and fit an internal parameter set that minimizes the error between the system simulation output and the system output through the system identification algorithm, and to use the internal parameter set to construct a building thermal fingerprint model.

[0051] In some implementations, the fingerprint modeling module transforms raw, discrete building operation monitoring data into a mathematical model capable of accurately reproducing the building's thermal dynamic response—a building thermal fingerprint model. The data preprocessing unit generates a time-synchronized and reliable standardized time-series dataset. This unit performs time alignment processing on the acquired historical operation dataset, unifying multiple data sources containing historical outdoor meteorological data sequences, historical indoor temperature data sequences, and historical equipment operating status data sequences to a fixed time step, such as 10 minutes, through interpolation or resampling methods to ensure a one-to-one correspondence between data points. Data cleaning operations are then performed using… The principle-based or local anomaly factor algorithm identifies and removes outliers caused by sensor malfunctions or communication anomalies, and fills in missing data points using linear interpolation or spline interpolation to ensure the integrity and continuity of the dataset. The variable definition unit sets clear input and output boundaries for the system identification algorithm. This unit uses the processed, standardized time-series dataset as the algorithm's input source. Outdoor dry-bulb temperature and total solar radiation intensity from historical outdoor meteorological data sequences, as well as air conditioning system start / stop status, actual output power, or variable-frequency compressor frequency from historical equipment operating status data sequences, are collectively defined as the system's external disturbances and control inputs. The historical indoor temperature data sequence is defined as the sole output of the system response. This step abstracts the complex building heat transfer process into a multi-input single-output (MISO) dynamic system problem, providing a clear mathematical framework for subsequent parameter fitting. The parameter fitting unit performs the core modeling task, solving for the internal parameter set that best describes the input-output relationship through iterative optimization. This unit employs system identification algorithms, such as prediction error method (PEM) or subspace identification, to construct a parameterized dynamic model structure, typically a state-space model or a transfer function model. This model contains a set of undetermined internal parameters. These parameters physically correspond to key thermal characteristics of the building envelope, such as equivalent thermal resistance and equivalent heat capacity. The algorithm fits the parameters by minimizing the error between the model's predicted output and the actual system output; its objective function is... It is usually expressed as the L2 norm of the prediction error, and its specific form is as follows:

[0052] ;

[0053] in, This represents the actual historical indoor temperature data sequence provided by the variable definition unit. This indicates that the building thermal fingerprint model is input into the same system. The simulated indoor temperature sequence was calculated. The parameter fitting unit repeatedly adjusted its internal parameter set using optimization algorithms such as gradient descent. This continues until the objective function J converges to its minimum value. This results in the optimized set of internal parameters. Once solidified, it forms a building thermal fingerprint model that can characterize the unique thermal performance of the existing building.

[0054] For example, the data preprocessing unit acquires raw monitoring data for a building over three consecutive time steps, with each step set to 10 minutes. The indoor temperature sensor data for the second step is missing due to a communication failure. This unit then processes the data from the preceding and following time steps... and Linear interpolation is performed on the measured values ​​to fill in the intermediate measured values. This generates a time-synchronized and complete standardized time-series dataset; the variable definition unit intervenes, defining the outdoor temperature at these three steps. ,like , , The frequency of the variable frequency compressor is defined as the control input quantity. The measured indoor temperature sequence Defined as the output quantity of the system response; the parameter fitting unit uses the prediction error method to fit the internal parameter set including equivalent thermal resistance and heat capacity. To perform fitting, the simulated sequence corresponding to the initial parameters is set as follows: According to the objective function formula The L2 norm of the prediction error was calculated. The internal parameter set was repeatedly adjusted through algorithm optimization. The values ​​in the table, when the parameters are optimized to their optimal values, correspond to the updated simulated sequence. At this point, the objective function value is recalculated. Since the error value has fallen below the preset minimum convergence threshold, the iteration stops and the final optimized internal parameter set is solidified into a unique building thermal fingerprint model for the building.

[0055] In one possible implementation, combining Figure 2 The co-simulation engine module includes:

[0056] The model update unit is used to update the physical property parameters in the building thermal fingerprint model according to the subset of the modification measures parameters, and generate a hypothetical building model after modification.

[0057] The logic loading unit is used to load the active environmental system control logic model into the hypothetical modified building model to form an integrated digital twin simulation.

[0058] The dynamic simulation unit is used to drive the integrated digital twin simulation body to run under typical meteorological data and record the cumulative system energy consumption and indoor temperature fluctuation data during the simulation process.

[0059] The index calculation unit is used to calculate the predicted energy consumption index based on the cumulative energy consumption value of the system and to calculate the predicted temperature stability index based on the indoor temperature fluctuation data, which together constitute the comprehensive performance evaluation result.

[0060] In some implementations, the co-simulation engine module provides quantitative performance predictions for each potential combination of renovation schemes. Its core lies in combining a static building model with dynamic control strategies to reproduce the entire operational process of the renovated building in a virtual environment. The model update unit transforms abstract renovation measure parameters into specific changes to the model's physical properties. This unit receives a subset of renovation measure parameters from the multi-objective optimization module, such as parameters including an external wall insulation layer thickness of 150 mm and an external window heat transfer coefficient of... Air tightness rating ACH50 is Specific numerical values ​​are obtained. These parameters are substituted into the state-space equations or transfer functions of the building thermal fingerprint model, directly modifying the coefficients representing the equivalent thermal resistance and equivalent heat capacity of the building envelope, thereby generating a hypothetical post-modification building model reflecting the effects of specific modification measures. The logic loading unit constructs a closed-loop simulation environment that can fully simulate the physical-control interaction. This unit loads the established active environmental system control logic model, including the variable frequency regulation algorithm based on indoor and outdoor temperature differences and load changes, and the intelligent defrosting trigger logic based on frost risk prediction, into the generated hypothetical post-modification building model. This loading process realizes bidirectional data exchange between models through a software interface. That is, the indoor temperature of the building model is used as the input of the control logic, while the equipment output power calculated by the control logic is used as the heat source input of the building model. Together, they constitute an integrated digital twin simulation, realizing deep coupling between the building physical entity and the equipment control strategy. The dynamic simulation unit drives the simulation in a standardized external environment to collect dynamic data reflecting its performance throughout the year. This unit places the integrated digital twin simulation entity under a typical annual meteorological dataset, such as TMY3 (Typical Meteorological Year 3), to perform continuous dynamic simulations. The simulation period typically covers a complete heating season or the whole year, with time steps set to 15 to 60 minutes. During the simulation, this unit records two key time-series data streams at high frequency: the instantaneous power consumption of the active environmental system at each time step, and the indoor air temperature at the same time step. The index calculation unit processes the raw data output by the dynamic simulation unit, aiming to transform it into key performance indicators for multi-objective optimization. This unit first processes the system energy consumption data, calculating the cumulative system energy consumption value by integrating the instantaneous power over the entire simulation period, and then normalizing it to obtain the predicted energy consumption index. The calculation formula is as follows:

[0061] ;

[0062] in, For the first The system average power over a time step The duration of the time step. Simultaneously, this unit analyzes indoor temperature fluctuation data, quantifying the stability of the indoor environment by calculating the standard deviation of the temperature series or the time integral exceeding the set comfort zone (e.g., 20-26°C), generating a predicted temperature stability index. These two indices together constitute a comprehensive performance evaluation result corresponding to a subset of parameters for the current renovation measures, and are output as feedback to the multi-objective optimization module to drive the next round of iterative optimization.

[0063] For example, when evaluating a specific combination of renovation schemes, the model update unit first receives data including information on an external wall insulation layer thickness of 150 mm and an external window heat transfer coefficient of... Air tightness rating ACH50 is A subset of parameters representing the modification measures with equivalent values ​​is used to modify the coefficients representing the equivalent thermal resistance and heat capacity of the building envelope in the building thermal fingerprint model, generating a hypothetical modified building model. Subsequently, the logic loading unit loads the active environmental system control logic model containing the frequency conversion regulation algorithm into this model, and establishes a feedback mechanism between indoor temperature input and equipment power output through a software interface, forming an integrated digital twin simulation. The dynamic simulation unit drives this simulation to run under typical annual meteorological data, setting the time step. for The system average power was recorded for two consecutive time steps over hours. and The final indicator calculation unit is based on the formula. The raw energy consumption data is processed and used to calculate predicted energy consumption indicators. And simultaneously analyze the indoor air temperature during that period, such as and Compared to setting a comfort zone The fluctuations in temperature are analyzed to generate a predicted temperature stability index, thereby completing a comprehensive performance evaluation of the current combination of modification schemes.

[0064] In one possible implementation, combining Figure 2 The database and control logic building module includes:

[0065] The frost data processing unit is used to acquire the historical operating dataset of the outdoor unit of the active environmental system, and process the dataset to generate a multi-dimensional environmental parameter sample containing frost event markers.

[0066] The prediction model training unit is used to train a frost event prediction model based on the multidimensional environmental parameter samples to predict the risk of frost based on real-time environmental parameters.

[0067] The control command integration unit is used to integrate the frost event prediction model into the active environmental system control logic model, so as to generate a pre-defrosting control command when the frost risk output by the frost event prediction model exceeds a preset trigger threshold.

[0068] In some implementations, the database and control logic building module develops a data-driven intelligent defrosting triggering strategy and integrates it into an executable active environmental system control logic model, replacing traditional passive defrosting methods based on timing or fixed temperature thresholds. The frosting data processing unit extracts and constructs a frosting feature sample library for machine learning from massive historical data. This unit first retrieves segments related to winter heating conditions from the historical operating data of the outdoor unit of the active environmental system. This data includes outdoor dry-bulb and wet-bulb temperatures, outdoor coil temperatures, compressor operating frequency, fan speed, and system power consumption. By analyzing these data sequences, the unit automatically identifies and marks actual frosting events. The criterion is usually a sustained temperature difference between the outdoor coil temperature and the outdoor wet-bulb temperature below a certain threshold, such as -3°C, accompanied by a decrease in system heating capacity. After event marking, the system uses data within a time window before the frosting event, such as 30 minutes, as a positive sample and data under normal frost-free operating conditions as a negative sample, thereby generating a multi-dimensional environmental parameter sample containing frosting event markers. The prediction model training unit trains a frost event prediction model based on the aforementioned sample database, capable of predicting future frost risk according to real-time operating conditions. This unit employs supervised learning algorithms, such as Support Vector Machine (SVM) or Gradient Boosting Decision Tree (GBDT), using multi-dimensional environmental parameter samples as the training set. The model uses the combination of real-time environmental parameters as the input feature vector. The output is a binary label indicating whether frost has occurred. The training process aims to find an optimal mapping function. This allows the model's predictions to fit the actual frosting labels to the greatest extent possible. The model's final output is not a simple "frost" or "no frosting" judgment, but a continuous value between 0 and 1, representing the probability of frosting risk. Its mathematical expression is:

[0069] ;

[0070] in, It represents the risk of frost formation and is a dimensionless probability value. It is a multi-dimensional feature vector that includes real-time measurements of outdoor ambient temperature, humidity, coil temperature, etc. This represents a fully trained frost event prediction model; Input the total number of features; For feature vectors The Middle These are normalized feature quantities, such as outdoor ambient temperature, relative humidity, coil temperature, and compressor frequency. These are the weight coefficients for the corresponding features obtained from model training; The bias constant is used. An iterative optimization algorithm minimizes the loss function to find the optimal weight vector and bias, enabling the model to accurately calculate the frosting risk probability of the current system given a combination of real-time environmental parameters. The control command integration unit executes the deployment task, embedding the trained prediction model into the actual control logic to achieve predictive maintenance. This unit will solidify the frosting event prediction model. As a functional module, it is integrated into the control logic model of the active environmental system. During system operation, the controller periodically collects real-time environmental parameters to form a feature vector. It then calls the model to calculate the current risk of frost formation. This risk value will be correlated with a preset trigger threshold. This threshold can be adjusted based on the balance requirements between energy efficiency and comfort, and is generally set between 0.85 and 0.95 for comparison. Once detected... Exceed The control logic no longer waits for the traditional defrosting conditions to be met, but immediately generates and issues pre-defrosting control commands to actively trigger the defrosting cycle, thereby avoiding a significant drop in system performance and a surge in energy consumption caused by severe frost.

[0071] For example, in the construction of the intelligent defrosting logic of an active environmental system, the prediction model training unit first obtains the normalized real-time environmental feature vector. Assuming outdoor temperature is included For example, a normalized value of 0.2 and relative humidity. , such as 0.9 and coil temperature For example, 0.1, and substitute it into the weight coefficients for training optimization. With bias According to the formula The calculation is performed, and its exponent term is: To obtain the probability of frost formation The control command integration unit compares the real-time risk value with the preset trigger threshold. The comparison shows that the current risk value has not exceeded the threshold, so we will continue to maintain efficient heating operation. However, if the humidity increases... Rise to It will immediately generate and issue a pre-defrost control command, and actively enter the defrost cycle before the heat exchange efficiency decreases.

[0072] In one possible implementation, combining Figure 2 The multi-objective optimization module includes:

[0073] Population initialization unit, used to initialize a population consisting of a subset of several modification measure parameters;

[0074] The fitness evaluation unit is used to call the co-simulation engine module to perform co-simulation steps for each subset of modification measure parameters in the population, obtain the corresponding comprehensive performance evaluation results, and calculate the fitness based on the comprehensive performance evaluation results.

[0075] An evolutionary operation unit is used to generate a new generation of population based on the fitness through selection, crossover, and mutation operations;

[0076] The iterative loop unit is used to repeatedly trigger the fitness evaluation unit and the evolutionary operation unit to execute the steps until the preset termination condition is met, and output the set of non-dominated solutions in the final generation population as the Pareto optimal solution set.

[0077] In some implementations, the multi-objective optimization module efficiently searches and identifies a series of modification schemes that achieve the optimal trade-off between the conflicting objectives of energy consumption and comfort within a vast solution space containing thousands of modification combinations. This module typically employs a population-evolutionary multi-objective optimization algorithm, such as Non-Dominated Sorting Genetic Algorithm II (NSGA-II). The population initialization unit serves as the initial starting point for generating the search algorithm, i.e., a diverse set of candidate schemes. It considers the parameter ranges set for each modification component in the modification measures database, such as the thickness of the external wall insulation layer between 50 mm and 250 mm, and the heat transfer coefficient of the external window within a certain range. to Between these steps, a series of subsets of modification measure parameters are randomly generated to form the initial population. The size of this population, for example, 100 individuals, directly affects the breadth of the search. The fitness evaluation unit provides a quantitative performance evaluation for each candidate solution in the population. For each subset of modification measure parameters in the population, this unit calls the co-simulation engine module, inputs the parameter subset, and performs a complete dynamic co-simulation. After the simulation is completed, the co-simulation engine module returns a comprehensive performance evaluation result including predicted energy consumption and predicted temperature stability indicators. In the context of multi-objective optimization, these two indicators themselves constitute the fitness of the solution, representing its coordinates in the objective space. The goal of the algorithm is to find a solution that can simultaneously minimize the values ​​of these two indicators. The evolutionary operation unit generates a next-generation population with better performance based on the fitness of the current population. This unit first selects individuals located at the current Pareto front or in sparse regions as parents through a selection operation, based on non-dominated sorting and crowding distance calculation. Next, a crossover operation is performed on the selected parents with a relatively high crossover probability, such as 0.9, that is, exchanging some of their parameters to generate new offspring solutions, thereby inheriting superior characteristics. Finally, a small random perturbation is applied to some parameters of the offspring individuals with a low mutation probability, such as 0.01 to 0.05, to introduce new possibilities and avoid the algorithm getting trapped in local optima. The iterative unit controls the entire optimization process, ensuring that the algorithm has enough time to converge to the global optimum. This unit repeatedly triggers the execution steps of the fitness evaluation unit and the evolutionary operation unit, allowing the population to evolve towards a better Pareto front generation after generation. This iterative process continues until a preset termination condition is met, such as reaching the maximum number of generations, such as 200 generations, or the Pareto front no longer showing significant improvement for several consecutive generations. When the termination condition is met, the algorithm selects all non-dominated solutions from the final generation of the population. These solutions do not have any solution that is completely superior to the other in terms of both energy consumption and comfort objectives. The set formed by these solutions is the final Pareto optimal solution set, providing a rich set of balanced alternatives for subsequent decisions. Figure 3 As shown, the distribution of the Pareto optimal solution set generated by the multi-objective optimization module in the objective space is illustrated, reflecting the trade-off between the two conflicting objectives of predicting energy consumption and temperature stability in different combinations of modification measures.

[0078] For example, when the multi-objective optimization module is searching for the best solution, the population initialization unit determines the thickness of the external wall insulation layer (50 to 250 mm) and the heat transfer coefficient of the external window based on the measures database. to Given the parameter range, an initial population of 100 individuals is randomly generated, where the parameter combination for individual A is "". Thick insulation layer "External window"; subsequently, the fitness evaluation unit calls the co-simulation engine to calculate the predicted energy consumption index of the individual. and predicted temperature stability index This is used as the target space coordinate; the evolutionary operation unit compares this individual with another individual B, with the parameter " Thick insulation layer "Exterior window" is the indicator. A Pareto dominance comparison was performed, determining that the two elements were not mutually dominant, meaning A had lower energy consumption but B was more stable, and this element was retained as the parent. Next, the unit swapped the parameter positions of the two elements with a crossover probability of 0.9 to generate a new offspring, and then made a small perturbation to the offspring's outer window parameters with a mutation probability of 0.02, such as adjusting to... To introduce new features, the cyclic iterative unit controls the population to repeat the evaluation and evolution process for 200 generations, and finally selects the set of non-dominated solutions that cannot dominate each other among the objectives, which is output as the Pareto optimal solution set.

[0079] In one possible implementation, combining Figure 2 The robustness stress testing module includes:

[0080] An extreme operating condition unit is used to place the integrated digital twin simulation corresponding to each scheme into an extreme climate test sequence for operation;

[0081] The failure event monitoring unit is used to monitor and record temperature exceedance events and defrost temperature drop events that occur in the integrated digital twin simulation under the extreme climate test sequence.

[0082] The reliability quantification unit is used to calculate and generate robustness evaluation indicators characterizing the reliability of the solution under extreme operating conditions based on the quantified data of the temperature exceedance event and the defrost temperature drop event.

[0083] In some implementations, the robustness stress testing module performs a deeper reliability check on the Pareto optimal solution set selected through multi-objective optimization. This aims to evaluate the performance of these solutions under extreme weather conditions exceeding normal design conditions, thereby identifying truly resilient and robust modification solutions. The extreme condition operation unit creates an extreme challenge environment for each candidate solution. This unit sequentially calls each solution in the Pareto optimal solution set and runs its corresponding integrated digital twin simulation within a predefined extreme climate test sequence. This sequence is not typical annual meteorological data, but rather represents historically rare but severe extreme weather events, such as a week-long period of extreme cold with outdoor temperatures consistently below [insert value here]. Or, for example, several consecutive days of blizzard weather, designed to simulate the maximum stress on a building's heating system. During extreme condition simulation operation, a failure event monitoring unit is activated simultaneously, aiming to capture and accurately record any critical events characterizing system performance degradation or failure in real time. This unit has two built-in monitoring logics. The first is for monitoring temperature exceedance events; it continuously monitors the indoor temperature in the simulation, and once the temperature falls below a preset minimum comfort or safety threshold, for example... The system immediately records the start time, duration, and lowest temperature of the event. The second system monitors defrost temperature drops. When the active environmental system control logic issues a defrost command, this unit pays special attention to changes in indoor temperature over a short period. If the temperature exceeds a specific threshold during defrosting, for example... A rapid drop in temperature is marked as a severe defrost temperature drop event and recorded. After the stress test simulations of all schemes are completed, the reliability quantification unit transforms the discrete, multi-dimensional failure events into a single, horizontally comparable robustness evaluation index. This unit summarizes the total duration of all temperature exceedance events and the total number of defrost temperature drop events for each scheme throughout the entire extreme climate test sequence. A weighted summation method is used to fuse these quantified failure data into a comprehensive index. Robustness Evaluation Index The calculation method can be expressed as:

[0084] ;

[0085] in, It refers to the duration of a single temperature exceeding the limit event. It is the total number of recorded defrosting temperature drop events, and and These are weighted coefficients based on the decision-maker's tolerance for temperature runaway and comfort fluctuations; both coefficients are dimensionless penalty factors. The robustness evaluation index is calculated using this formula. The smaller the value, the more reliable and robust the solution is under extreme conditions. This indicator, along with the predicted energy consumption and predicted temperature stability indicators, will be submitted to the solution decision and output module as an important basis for the final decision. Figure 4 As shown, the real-time indoor temperature change curves of candidate schemes in the Pareto optimal scheme set under extreme climate test sequences are displayed, intuitively demonstrating the ability of robust schemes to maintain indoor temperature above the safe threshold under extreme conditions.

[0086] For example, when the robustness stress testing module performs reliability verification on the Pareto optimal solution set, the extreme condition operation unit first places the integrated digital twin simulation of a specific modification scheme into an extreme cold wave weather sequence containing 48 consecutive hours of outdoor temperatures below -20°C for simulation operation; during the simulation, the failure event monitoring unit monitors the indoor temperature in real time and records that the scheme experienced two instances of temperatures below -20°C under this sequence. The durations of the events exceeding the safety thresholds were 1.2 hours and 0.8 hours, respectively. Simultaneously, it was observed that three temperature drops exceeding [a certain value] occurred under extremely low temperature defrosting conditions. A severe defrosting temperature drop event, namely Ultimately, the reliability quantification unit is based on the weighting coefficients set by the decision-maker. and Substitute into the formula Calculations were performed to derive the robustness evaluation index of the scheme. The smaller the quantitative score, the higher the system resilience and operational robustness of the solution under extreme pressure conditions.

[0087] In one possible implementation, combining Figure 2 The indicator calculation unit includes:

[0088] A deviation extraction subunit is used to extract a temperature deviation sequence from the indoor temperature fluctuation data;

[0089] The sub-index calculation sub-unit is used to calculate the standard deviation of the temperature deviation sequence as the first stability sub-index, and to calculate the percentage of time the indoor temperature exceeds the preset range as the second stability sub-index.

[0090] The weighted fusion subunit is used to perform weighted fusion of the first stability sub-index and the second stability sub-index to generate a predicted temperature stability index.

[0091] In some implementations, the index calculation unit transforms the raw indoor temperature fluctuation data generated during dynamic simulation into a comprehensive index that can fully and accurately quantify the stability of the indoor thermal environment. The deviation extraction subunit aims to separate the actual indoor temperature fluctuations from the ideal setpoint. This subunit reads the time series of indoor temperature fluctuation data generated by the dynamic simulation unit and, based on a preset indoor temperature control target value, such as the winter heating setpoint... The temperature deviation at each time step is calculated, thus generating a temperature deviation sequence reflecting the degree of deviation between the actual temperature and the target value. The sub-index calculation unit characterizes temperature stability from two different dimensions. First, this sub-unit calculates the standard deviation of the temperature deviation sequence generated in the previous step, obtaining the first stability sub-index. This indicator directly reflects the dispersion or fluctuation range of the indoor temperature around the setpoint throughout the simulation cycle; the smaller the value, the more stable the temperature regulation by the control system. Secondly, this subunit is based on a preset indoor temperature comfort range, such as the range recommended by the ASHRAE 55 standard. to The system calculates the cumulative duration of indoor temperatures exceeding the specified range and the proportion of this duration to the total simulation duration, generating a second stability sub-index. This indicator quantifies the frequency and duration of users experiencing uncomfortable environments; the smaller the value, the higher the compliance rate of the indoor thermal environment. The weighted fusion sub-unit performs the task of synthesizing the indicator. Its engineering goal is to merge two sub-indicators into a single, easily optimizable predictive temperature stability indicator based on the importance of different evaluation dimensions. To ensure the reasonable merging of indices with different physical dimensions, this unit first performs dimensionless processing on the sub-indices. The final predicted temperature stability index is calculated using the following weighted fusion formula:

[0092] ;

[0093] in, The calculated temperature standard deviation is the first stability sub-index. The calculated percentage of time the temperature exceeded the limit, i.e., the second stability sub-index; For example, set to a reference temperature standard deviation. , used to Dimensionless; and The weighting coefficients assigned to the two dimensionless sub-indicators are set according to the different emphases of the project on temperature fluctuation sensitivity and comfort compliance rate, and their sum is 1. The predicted temperature stability index obtained through this formula... The smaller the value, the more stable the indoor thermal environment provided by the renovation plan. This indicator, together with the predicted energy consumption indicator, will constitute the final comprehensive performance evaluation result.

[0094] For example, when processing the indoor temperature stability evaluation, the index calculation unit first reads the indoor temperature sequence generated by the simulation by the deviation extraction subunit, such as five consecutive time steps with a time interval of... Target values ​​are set based on winter heating. The corresponding deviation sequence is calculated as follows: Subsequently, the sub-index calculation unit statistically analyzed the deviation series and calculated the standard deviation to be approximately... That is, the first stability sub-index At the same time, in accordance with the recommendations of ASHRAE 55 standard to The comfort zone is determined by the sequence being within a comfortable range. However, assuming the total duration exceeding the comfort zone is 24 hours and the total simulation duration is 240 hours over a longer simulation period, the calculated duration percentage is 0.10, which is the second stability sub-index. Finally, the weighted fusion sub-unit is based on the set weight coefficients. and benchmark standard deviation Substitute into the formula Calculations were performed to obtain the final predicted temperature stability index. The smaller this value, the more stable the indoor thermal environment, and together with the energy consumption index, it constitutes the comprehensive evaluation result.

[0095] In one possible implementation, combining Figure 2 The scheme decision and output module includes:

[0096] The matrix construction unit is used to standardize the predicted energy consumption index, predicted temperature stability index, and robustness evaluation index of each scheme in the Pareto optimal scheme set to generate a standardized performance matrix.

[0097] The weighted scoring unit is used to assign weight coefficients to each performance index in the standardized performance matrix and to calculate the comprehensive score of each scheme.

[0098] The scheme selection unit is used to select the scheme with the highest comprehensive score as the final modification scheme.

[0099] In some implementations, the scheme decision and output module transforms the mathematically equivalent Pareto optimal scheme set generated by multi-objective optimization into a unique and executable final modification scheme by incorporating the decision-maker's subjective preferences and risk considerations. The matrix construction unit eliminates the incomparability between different performance indicators caused by differences in dimensions and numerical ranges. This unit receives three core performance indicators for each scheme in the Pareto optimal scheme set: predicted energy consumption, predicted temperature stability, and robustness evaluation. Since these three indicators are all cost-related, smaller values ​​are better. This unit uses a min-max normalization method to process them, mapping them to the interval between 0 and 1. The calculation formula is as follows:

[0100] ;

[0101] in, Let be the standardized value of a certain indicator of any given scheme. This is the original value of the indicator, while and These represent the maximum and minimum values ​​of the indicator within the entire Pareto optimal set. This operation generates a standardized performance matrix, where each row represents a modification scheme and each column represents a dimensionless performance score. The weighted scoring unit quantifies the decision-maker's preferences and incorporates them into the evaluation process. This unit assigns corresponding weight coefficients to each performance indicator in the standardized performance matrix. For example, if the project prioritizes energy efficiency, the weight coefficient for the predicted energy consumption indicator might be set to 0.5; if operational reliability under extreme weather conditions is crucial, the weight for the robustness evaluation indicator might be increased to 0.4. These weight coefficients are jointly agreed upon by the project decision-making team based on factors such as investment budget, usage requirements, and risk preferences, and the sum of all weight coefficients must be 1. Next, this unit calculates a comprehensive score for each scheme, which is calculated as a weighted sum of the standardized performance scores:

[0102] ;

[0103] in, This is the overall score of the plan; , , These are the weighting coefficients assigned to the three indicators: predicted energy consumption, temperature stability, and robustness. , , This refers to the standardized value of the corresponding indicator. The solution selection unit performs the final decision-making step, selecting the optimal solution based on the quantified comprehensive score. This unit's comprehensive score for all solutions... The solutions are sorted, and the one with the highest overall score is selected. This highest-scoring solution, which is the globally optimal solution after fully considering energy consumption, comfort, and reliability under extreme conditions, and taking into account the decision-maker's specific preferences, will be determined as the final renovation plan and output by this module to guide subsequent design and construction.

[0104] For example, when determining the final renovation plan, the scheme decision and output module first uses the matrix construction unit to summarize the indicators of each scheme in the Pareto optimal scheme set. Taking the predicted energy consumption indicator as an example, if the maximum value in the set... for minimum value for For the original value of a certain scheme A for Substitute into the formula Standardized score is calculated Subsequently, the weighted scoring unit assigns weight coefficients based on item preferences. , , If the three standardized scores of scheme A are respectively Substitute into the formula Calculate the overall score Finally, the scheme selection unit sorts all schemes in descending order of their comprehensive scores. If scheme A's score is... If the value is the highest in the entire set, then it is determined as the final modification scheme and output.

[0105] In one possible implementation, combining Figure 2 The scheme decision and output module further includes:

[0106] The model-driven update unit is used to import the attributes of the three-dimensional component into the associated building information model platform to update the physical attribute parameters of the corresponding model component in the platform.

[0107] The configuration file generation unit is used to encapsulate the system control parameters into a configuration file that can be read by the controller and send it to the associated active environmental system controller.

[0108] In some implementations, the scheme decision and output module, after determining the final renovation scheme, achieves automated and high-fidelity data flow from optimization decision-making to engineering practice, ensuring that the analysis results can be seamlessly applied to the design and control levels. The model-driven update unit directly synchronizes the building physics renovation portion of the decision scheme to the digital design model. This unit receives the three-dimensional component attributes included in the final renovation scheme, such as the insulation layer material type and thickness of specific exterior wall components, and the heat transfer coefficient and solar heat gain coefficient of exterior window components. Specific three-dimensional component attributes include: attributes of interior insulation components based on composite structures, with a 5mm thick polymer nano-insulation material at the bottom and an insulation layer superimposed on the interior side; and attributes of low-energy-consumption, high-performance energy-saving window components with optimized heat transfer coefficients. This unit connects to the associated Building Information Modeling (BIM) platform via an Application Programming Interface (API). Based on the unique identifiers (GUIDs) of the components, it can accurately locate the walls, windows, and other objects requiring modification within the BIM model and programmatically update the physical property parameters of these components. This ensures that the BIM model, as the single data source for the project, accurately reflects the final decision and provides a basis for subsequent detailed construction drawings and bill of materials generation. Simultaneously, the configuration file generation unit processes the system control portion of the decision-making scheme, aiming to translate optimized control strategies into a set of instructions that the equipment can directly execute. This unit extracts system control parameters from the final modification scheme, including key inflection points on the variable frequency drive curve set for the active environmental system, frost risk thresholds in the intelligent defrosting trigger strategy (e.g., 0.9), and other refined operating parameters. The system control portion is associated with an air-source heat pump all-in-one fresh air system, which integrates winter heating, summer cooling, and active humidification functions. System control parameters include the cooling / heating switching thresholds set for this air-source heat pump all-in-one fresh air system, dynamic humidification adjustment logic, and key inflection points on the variable frequency drive curve. Subsequently, the unit encapsulates these numerical and logical parameters into a standardized configuration file, such as JSON or XML, according to the specific controller's protocol specifications. This configuration file is treated as a complete operating "recipe" and sent to the associated active environmental system controller via industrial Ethernet or building automation network protocols, such as BACnet / IP or Modbus TCP. After the controller receives and loads the configuration file, its internal control logic and operating setpoints are automatically updated to the optimized state, achieving seamless integration from simulation optimization to actual system debugging and maximizing the realization of the expected energy-saving effect. This invention constructs an ultra-low energy consumption building renovation system by logically coupling and synergistically optimizing three measures: high-performance energy-saving window renovation, composite insulation system renovation with 5mm polymer nanomaterial insulation layers on the ground floor and interior sides, and air-source heat pump fresh air system renovation with cooling / heating / humidification functions.Not only has the building's basic load been minimized through passive nano-composite insulation and high-performance windows, but the indoor environment has also been precisely regulated through an active air energy fresh air system. Ultimately, on the premise of ensuring the comfort of constant temperature, constant humidity, and constant oxygen indoors, a leapfrog improvement in building energy efficiency has been achieved. Through this way of collaborative optimization of software and hardware, the technical pain point of the difficulty in balancing energy-saving goals and comfort goals during the renovation of existing buildings has been solved.

[0109] Exemplarily, when the solution decision and output module executes the implementation of the final solution, the model-driven update unit accesses the building information model platform through the API interface. According to the unique identifier of the exterior window component determined in the final renovation plan, such as the GUID being "W-001", the heat transfer coefficient attribute of this component is changed from the original programmed update to , and geometric parameters such as the associated material thickness are synchronously updated; at the same time, the configuration file generation unit extracts the control parameters of the active system supporting this plan. For example, the frosting risk threshold in the intelligent defrost trigger strategy is set to , and the variable frequency adjustment command is calculated, and the temperature difference number at a certain time step and the coefficient obtained through optimization are brought in for calculation to obtain the recommended frequency of the compressor; this unit then encapsulates the above-mentioned threshold , frequency parameters, PID coefficients and other values into a standardized JSON configuration file and remotely issues it to the field controller through the Modbus TCP protocol. After the controller loads this file, it automatically updates the running logic to the optimized state.

[0110] It should be noted that the electrical connections between the above-mentioned various units do not necessarily mean direct connections of the lines. Indirect connection methods, as long as they achieve the purpose of the present invention, are applicable to the embodiments of the present invention. The above are only exemplary embodiments of the present invention and cannot be used to limit the scope of the present invention.

[0111] That is, any equivalent changes and modifications made in accordance with the teachings of the present invention still fall within the scope covered by the present invention. Those skilled in the art will easily think of other implementation schemes of the present invention after considering the specification and the disclosure of the practical truth. This application aims to cover any variations, uses or adaptation changes of the present invention, and these variations, uses or adaptation changes follow the general principles of the present invention and include the common knowledge or conventional technical means in the technical field not recorded in the present invention.

Claims

1. An energy-saving system for converting existing buildings into ultra-low energy consumption buildings, characterized in that, The system includes: The data acquisition module is used to acquire historical operation datasets that include historical outdoor meteorological data sequences, historical indoor temperature data sequences, and historical equipment operation status data sequences. The fingerprint modeling module is used to construct a building thermal fingerprint model that represents the actual thermal dynamic response of a building based on the historical operating dataset and through a system identification algorithm. The database and control logic building module is used to establish a database of renovation measures containing several energy-saving renovation components, and to establish an active environmental system control logic model containing frequency conversion regulation strategy and intelligent defrosting trigger strategy. The collaborative simulation engine module is used to select a subset of retrofit measure parameters from the retrofit measure database, couple the building thermal fingerprint model, the subset of retrofit measure parameters, and the active environmental system control logic model, run collaborative simulation, and generate a comprehensive performance evaluation result corresponding to the subset of retrofit measure parameters. The multi-objective optimization module is used to employ a multi-objective optimization algorithm to iteratively optimize different subsets of modification measure parameters, with the goal of optimizing the predicted energy consumption index and predicted temperature stability index in the comprehensive performance evaluation results, and to select a Pareto optimal solution set. The robustness stress test module is used to acquire extreme climate test sequences for simulating extreme weather, and to perform robustness stress tests on each scheme in the Pareto optimal scheme set to obtain robustness evaluation indicators for each scheme. The scheme decision and output module is used to determine the final modification scheme from the Pareto optimal scheme set based on the comprehensive performance evaluation results and the robustness evaluation index, and output the modification decision data corresponding to the final modification scheme, which includes three-dimensional component attributes and system control parameters.

2. The energy-saving system for converting existing buildings into ultra-low energy consumption buildings according to claim 1, characterized in that, The fingerprint modeling module includes: The data preprocessing unit is used to perform time alignment and data cleaning on the historical running dataset to generate a standardized time-series dataset. The variable definition unit is used to input the standardized time series dataset into the system identification algorithm, using the historical outdoor meteorological data sequence and the historical equipment operating status data sequence as system inputs and the historical indoor temperature data sequence as system outputs. The parameter fitting unit is used to iteratively calculate and fit an internal parameter set that minimizes the error between the system simulation output and the system output through the system identification algorithm, and to use the internal parameter set to construct a building thermal fingerprint model.

3. The energy-saving system for converting existing buildings into ultra-low energy consumption buildings according to claim 1, characterized in that, The co-simulation engine module includes: The model update unit is used to update the physical property parameters in the building thermal fingerprint model according to the subset of the modification measures parameters, and generate a hypothetical building model after modification. The logic loading unit is used to load the active environmental system control logic model into the hypothetical modified building model to form an integrated digital twin simulation. The dynamic simulation unit is used to drive the integrated digital twin simulation body to run under typical meteorological data and record the cumulative system energy consumption and indoor temperature fluctuation data during the simulation process. The index calculation unit is used to calculate the predicted energy consumption index based on the cumulative energy consumption value of the system and to calculate the predicted temperature stability index based on the indoor temperature fluctuation data, which together constitute the comprehensive performance evaluation result.

4. The energy-saving system for converting existing buildings into ultra-low energy consumption buildings according to claim 1, characterized in that, The database and control logic construction module includes: The frost data processing unit is used to acquire the historical operating dataset of the outdoor unit of the active environmental system, and process the dataset to generate a multi-dimensional environmental parameter sample containing frost event markers. The prediction model training unit is used to train a frost event prediction model based on the multidimensional environmental parameter samples to predict the risk of frost based on real-time environmental parameters. The control command integration unit is used to integrate the frost event prediction model into the active environmental system control logic model, so as to generate a pre-defrosting control command when the frost risk output by the frost event prediction model exceeds a preset trigger threshold.

5. The energy-saving system for converting existing buildings into ultra-low energy consumption buildings according to claim 1, characterized in that, The multi-objective optimization module includes: Population initialization unit, used to initialize a population consisting of a subset of several modification measure parameters; The fitness evaluation unit is used to call the co-simulation engine module to perform co-simulation steps for each subset of modification measure parameters in the population, obtain the corresponding comprehensive performance evaluation results, and calculate the fitness based on the comprehensive performance evaluation results. An evolutionary operation unit is used to generate a new generation of population based on the fitness through selection, crossover, and mutation operations; The iterative loop unit is used to repeatedly trigger the fitness evaluation unit and the evolutionary operation unit to execute the steps until the preset termination condition is met, and output the set of non-dominated solutions in the final generation population as the Pareto optimal solution set.

6. The energy-saving system for converting existing buildings into ultra-low energy consumption buildings according to claim 3, characterized in that, The robust stress testing module includes: An extreme operating condition unit is used to place the integrated digital twin simulation corresponding to each scheme into an extreme climate test sequence for operation; The failure event monitoring unit is used to monitor and record temperature exceedance events and defrost temperature drop events that occur in the integrated digital twin simulation under the extreme climate test sequence. The reliability quantification unit is used to calculate and generate robustness evaluation indicators characterizing the reliability of the solution under extreme operating conditions based on the quantified data of the temperature exceedance event and the defrost temperature drop event.

7. The energy-saving system for converting existing buildings into ultra-low energy consumption buildings according to claim 3, characterized in that, The indicator calculation unit includes: A deviation extraction subunit is used to extract a temperature deviation sequence from the indoor temperature fluctuation data; The sub-index calculation sub-unit is used to calculate the standard deviation of the temperature deviation sequence as the first stability sub-index, and to calculate the percentage of time the indoor temperature exceeds the preset range as the second stability sub-index. The weighted fusion subunit is used to perform weighted fusion of the first stability sub-index and the second stability sub-index to generate a predicted temperature stability index.

8. The energy-saving system for converting existing buildings into ultra-low energy consumption buildings according to claim 1, characterized in that, The scheme decision and output module includes: The matrix construction unit is used to standardize the predicted energy consumption index, predicted temperature stability index, and robustness evaluation index of each scheme in the Pareto optimal scheme set to generate a standardized performance matrix. The weighted scoring unit is used to assign weight coefficients to each performance index in the standardized performance matrix and to calculate the comprehensive score of each scheme. The scheme selection unit is used to select the scheme with the highest comprehensive score as the final modification scheme.

9. The energy-saving system for converting existing buildings into ultra-low energy consumption buildings according to claim 1, characterized in that, The scheme decision and output module also includes: The model-driven update unit is used to import the attributes of the three-dimensional component into the associated building information model platform to update the physical attribute parameters of the corresponding model component in the platform. The configuration file generation unit is used to encapsulate the system control parameters into a configuration file that can be read by the controller and send it to the associated active environmental system controller.

10. An energy-saving method for converting existing buildings into ultra-low energy consumption buildings, characterized in that, The method is used in an energy-saving system for converting an existing building into an ultra-low energy consumption building as described in any one of claims 1-9, the method comprising: Obtain a historical operational dataset containing historical outdoor meteorological data sequences, historical indoor temperature data sequences, and historical equipment operating status data sequences; Based on the historical operating dataset, a building thermal fingerprint model representing the real thermal dynamic response of a building is constructed using a system identification algorithm. Establish a database of retrofit measures containing several types of energy-saving retrofit components, and establish an active environmental system control logic model that includes frequency conversion regulation strategy and intelligent defrosting trigger strategy; A subset of retrofit measure parameters is selected from the retrofit measure database, and the building thermal fingerprint model, the subset of retrofit measure parameters, and the active environmental system control logic model are coupled together. Co-simulation is then performed to generate a comprehensive performance evaluation result corresponding to the subset of retrofit measure parameters. A multi-objective optimization algorithm is adopted to optimize the predicted energy consumption index and predicted temperature stability index in the comprehensive performance evaluation results. Different subsets of modification measures are iteratively optimized to select the Pareto optimal solution set. Obtain extreme climate test sequences for simulating extreme weather, conduct robust stress tests on each scheme in the Pareto optimal scheme set, and obtain robustness evaluation indicators for each scheme. Based on the comprehensive performance evaluation results and the robustness evaluation index, the final modification scheme is determined from the Pareto optimal scheme set, and the modification decision data corresponding to the final modification scheme, including three-dimensional component attributes and system control parameters, is output.

Images